• Written by Azhar Aliyar, Director

In practice, AsscherAi often fails to provide useful answers about an organisation's operations because it lacks access to the right data, specifically the process and exception records, project status, budget consumption, and supplier performance data that operations teams rely on. This is not because the AI is flawed, but because the necessary data is not available or is not in a usable form.

The systems that operations depend on

The purchase order table, project management system, and maintenance log are just a few examples of the systems that operations teams use to track and manage their work. These systems contain critical data that AsscherAi needs to provide accurate and useful answers. For instance, a shift supervisor checks the maintenance log to see what equipment is due for maintenance, and then uses this information to plan the day's work. Without access to this data, AsscherAi would not be able to answer questions about equipment maintenance or provide insights into operational efficiency.

A category manager, for example, compares two suppliers based on their performance data, which is stored in the supplier performance system. This data includes information on delivery times, product quality, and pricing. If AsscherAi does not have access to this data, it will not be able to answer questions about supplier performance or provide recommendations for improvement.

What a missing system does to an answer

If a critical system is missing, AsscherAi's answers will be incomplete or inaccurate. For example, if the budget consumption data is not available, AsscherAi will not be able to answer questions about project costs or provide insights into budget management. This can lead to poor decision-making and a lack of trust in the AI system. Most people assume that AsscherAi can somehow magically fill in the gaps, but this is wrong: the AI is only as good as the data it has access to.

A missing system can also affect the accuracy of AsscherAi's answers in subtle ways. For instance, if the project status system is not available, AsscherAi may not be able to take into account the current status of a project when answering questions about it. This can lead to answers that are out of date or irrelevant, which can be frustrating for users and undermine their confidence in the system.

The definitions that must be agreed

For AsscherAi to provide accurate and useful answers, there must be a shared understanding of key terms and definitions. This includes definitions of operational metrics, such as uptime and downtime, as well as definitions of project status and budget categories. If these definitions are not agreed upon, AsscherAi's answers may be inconsistent or misleading. A clear understanding of operations is essential for effective use of the AI system.

Agreeing on definitions can be a time-consuming process, but it is essential for ensuring that AsscherAi provides accurate and useful answers. It requires close collaboration between operations teams, data analysts, and other stakeholders to ensure that everyone is using the same language and metrics.

Quality problems that only surface later

Quality problems with the data can be difficult to detect, especially if they are subtle or only affect certain types of queries. For example, if the data in the maintenance log is not up to date, AsscherAi's answers about equipment maintenance may be inaccurate. These problems can only be detected by carefully reviewing the data and testing the AI system with a variety of queries.

Quality problems can also arise from the way the data is structured or formatted. For instance, if the data in the supplier performance system is not in a standard format, AsscherAi may have difficulty parsing it or using it to answer questions. This can lead to errors or inconsistencies in the AI's answers, which can be frustrating for users and undermine their confidence in the system.

When AsscherAi is the wrong choice

AsscherAi is not a replacement for human judgment or expertise. There are many situations in which it is better to consult with a human expert or use a different tool or system. For example, if a complex operational problem requires a nuanced or creative solution, AsscherAi may not be the best choice. In these situations, it is better to contact us to discuss the problem and determine the best approach.